Demonstration of speech enhancement in highly non-stationary multisource environment (CHiME Challenge)

Demonstration of the results obtained for the PASCAL 'CHiME' Challenge

(Speech enhancement in highly non-stationary, multisource noisy environment)

Here we have sound demonstration of speech enhancement in highly non-stationary
multisource noisy environment. The original noisy data are available at the CHiME Challenge homepage.
and the speech enhancement was done at NTT Communication Science Laboratories.
The related references can be found at the bottom of this page.
The sound volumes of the demo sounds are normalized by their maximum absolute value.

To proceed you need the Quicktime pluggin.
The pages have been tested with Firefox 8.0 and Internet Explorer 6.0

1. Development set

Female speaker
Male speaker

2. Evaluation set

Female speaker
Male speaker

References

1. Delcroix, M., Kinoshita, K., Nakatani, T., Araki, S., Ogawa, A., Hori, T., Watanabe, S., Fujimoto, M., Yoshioka, T., Oba, T., Kubo, Y., Souden, M., Hahm S.-J. and Nakamura, A., ``Speech Recognition in the Presence of Highly Non-Stationary Noise Based on Spatial, Spectral and Temporal Speech/Noise Modeling Combined with Dynamic Variance Adaptation,'' in Proc. CHiME Workshop, pp. 12-17, 2011.

2. Nakatani, T., Araki, S., Delcroix, M., Yoshioka, T. and Fujimoto, M., ``Reduction of highly nonstationary ambient noise based on spectral and locational characteristics of speech and noise for robust ASR,'' Proc. Interspeech'11, pp. 1785-1788, 2011.

3. Kinoshita, K., Souden, M., Delcroix, M. and Nakatani, T., “Single channel dereverberation using example-based speech enhancement with uncertainty decoding technique,’’ Proc. Interspeech'11, pp. 197-200, 2011.